from the increase for tobacco programs and research designed to reduce .... newsletter contained articles, tables, and graphs detailing information about ...
John P Elder, MPH PhD Christine C. Edwards, MPH Terry L. Conway, PhD Erin Kenney, PhD C. Anderson Johnson, PhD Erica D. Bennett, MPH Dr. Elder, Ms. Edwards, Dr. Conway, and Dr. Kenney are with San Diego State University. Dr. Elder is a Professor and Head of the Health Promotion Division at the Graduate School of Public Health, Ms. Edwards is Project Director for the Independent Evaluation Project, Dr. Conway is Research Director for the Independent Evaluation Project and lecturer in the Graduate School of Public Health, and Dr. Kenney is the former Project Director for the Independent Evaluation Project. Ms. Bennett was Evaluation Coordinator for the Independent Evaluation Project at the time of the study. Dr. Johnson is Principal Investigator of the Tobacco Industry Response Evaluation and the Sidney Garfield Professor of Preventive Medicine at the University of Southern California.
Address correspondence to Dr. Elder, 9245 Sky Park Ct., Suite 221, San Diego, ca 92123; tel. 619-594-2997;fax 619-5942998; e-mail
July/August 1 996 * Volume I I I
Independent Evaluation of the California Tobacco Education Program
SYNOPSIS
Objective. To monitor the implementation of tobacco control programs and research in accordance with Califomia's Proposition 99, approved by the voters in 1988, which increased the state's cigarette tax by 25 cents and designated one quarter of the increased revenue-approximately $100 million per year-to develop statewide media campaigns and to fund local heafth departments, community-based organizations, schools, and agencies working with high risk populations. Methods. The authors evaluated the extent and the effectiveness of the implementation independently with standardized forms developed to track any tobacco-related information and activities, local media campaigns, changes in local policies and ordinances, training sessions, and prevention and cessation programs. Activities were reported on the forms to the authors monthly or quarterly. Results. The evaluation indicated that local heakh departments, communitybased organizations, and other groups produced an exceedingly high volume of diverse tobacco control activities throughout the state. They reached a variety of ethnic groups and high risk populations with their programs. The data also show that local health departments and competitive grantees responded with an overall shift in their approach to tobacco control and moved from interventions focused primarily on individual people to broader, more community- and environment-oriented interventions. Additionally, analysis of the wholesale tobacco tax revenues revealed that per capita purchasing continued to decline in California at a rate greater than in the rest of the United States. Conclusion. The evaluation demonstrated that Proposition 99 accomplished much of what it set out to do-reducing tobacco prevalence, reaching out to underserved populations, and heightening the awareness of the dangers of environmental tobacco smoke.
Iw n 1988, the voters of California approved Proposition 99, a popularly initiated referendum that called for a 25-cent increase in the state cigarette tax and designated one-fourth of the $600 million annual revenue from the increase for tobacco programs and research designed to reduce the prevalence of smoking. This launched not only the most intensive tobacco control intervention but also one of the most aggressive public health interventions ever undertaken. Public Health
Reports
353
Scientific Contribution The enabling legislation enacted to implement Proposition 99 set aside approximately $100 million a year from the tax increase to develop statewide anti-tobacco media campaigns and fund local health departments, communitybased organizations, schools, and agencies working with high risk populations, such as ethnic minorities and pregnant women, to promote _ tobacco prevention and cessation.' To evaluate the extent and effectiveness of the Proposition 99 efforts, the State Department of Health Services contracted with the University of California-San Diego (UCSD) to conduct a bi-annual statewide telephone survey of adults and young people.2 San Diego State University (SDSU) was contracted to monitor and track independently the number and nature of programs implemented across the state.3 This paper focuses on the overall impact of Proposition 99 as well as the structure and findings of the SDSU process evaluation from 1990 through 1994.
Methods Impact evaluation. Although the independent evaluation was not responsible for overall incidence and prevalence surveys, evaluators were given the task of measuring adult per capita purchasing of cigarettes from published records of excise tax revenues collected by state governments and published by tobacco trade journals. Beginning with the year 1980, data were collected for California and the rest of the United States and were analyzed through examination of trends, of percent changes, and of regression procedures. Process evaluation activities. The first focus of the evaluation was on centrally funded institutional programs run through the local health departments designated as Local Lead Agencies (LLAs). Secondly, the evaluation examined community organizations funded through a competitive grant (CG) process. The third look was at naturally occurring "secular events," so labeled because they were not funded through the statewide initiative and may or may not have been stimulated by the campaign. For example, we conducted annual statewide surveys of physicians and dentists to monitor prevention and cessation activities among prnmary care providers. Another aspect of the evaluation was the monitoring of the tobacco industry's advertising and promotional efforts to derail Proposition 99 initiatives. 354 Public Health
Reports
A final element in the evaluation was the tracking of the efforts between and among communities set up in regional networks.
Data collection. The initial data collection effort in the evaluation process centered on the variables mandated by the original legislation and those developed jointly by staff members working on the SDSU independent eval,i 6uation, staff members of the a State Department of Health Services Tobacco Control *i *Section (TCS), and representatives from LLAs. The LLAs agreed to track tobacco-related activities, local media coverage, changes in local tobacco policies and _ ; I
ordinances, training sessions, and prevention and cessation programs. At the same time, the LLAs themselves were surveyed annually by the SDSU evaluators to gather more detailed information on their activities. The Competitive Grantee (CG) evaluation was similar to the LLA evaluation in the use of standardized forms to collect information. Activities were reported to SDSU on a monthly basis. All LLAs and CGs participated in the data collection. It was not possible, however, to estimate missing data rates, because the forms were submitted only if a particular activity or program was performed in a given month. The standardized evaluation forms for LLAs and CGs were revised first in 1992 based on requests from the field and on evidence that certain variables were not useful or valid. The forms were revised a second time in April 1994 at the request of the state to lessen the burden of data collection and reporting. Because of this ongoing revision process, it was impossible to assess some trends because of variables being dropped. A separate set of standardized data collection forms were developed to record activities funded by the state in 10 regions that tied geographic areas together with the common goal of promoting tobacco control (Figure 1). The regional forms were used on a quarterly basis, beginning in May 1992. As with the LLAs and CGs, all 10 geographic regions participated in this process evaluation. Evaluating the response of the tobacco industry to this massive effort was considered critical. The monitoring was done by the University of Southern California (USC), whose investigators assessed changes in marketing and promotion through the analyses of magazine, newspaper, and billboard advertising, as well as in-store promotions. In 1994, an additional component was added to track tobacco industry promotions like sporting and cultural events. July/August
1996 * Volume Ill
California's Tobacco Education
Results
Figure 1. California tobacco control regions North Coast Sierra Casade
These results focus on data reported from October 1990 through March
Del Norte,
1994.
Butt Colusa, Glenn, Lass[en, Modoc, Plumas, Shasta, Sierra, Sisidyou, Tehama, Thnlty, Yuba
Humboldt, Lake, Mendocino, Napa, Sonoma I'l
Al.vneda,Corwn
-
Francisco, San
Gold Country A'Inino' AnmAsnr- rsl2vgbr lnyo, Mono, Nemvada, Placer, SacrAnmento San Jloquin, Suizer, Tuolunnne, Yolo
~~~~Dorado, ~~~~~Stanislaus,
Santa Clara, Mateo, Costa \Marpsn,
Impact evaluation of tobacco purchasing. Through the last quarter of 1994, California's decline in wholesale cigarette purchases, adjusted per capita, continued at a rate faster than the rest of the United States. This was in spite
of (a) sales of major brands of cigarettes declining from late 1992 to mid1993 because of the introduction of generic cigarettes, and (b) patterns of immigration and out-migration, reBenito, Santa Cruz sulting in California's transition to iTn-County lower average levels of education and Tri-County__ income and, hence, a greater potential Imperal, San Luis Obispo, Riverside, for increased smoking.4Sata Barbara, San Bernardino Figure 2 shows the decline in Ventura Los AnphMs_ wholesale purchasing, adjusted per Los Angeles, city of capita, in the rest of the United States Long Beach, city of and in California. Purchasing levels Pisadena So rnCoast to diverge in the early 1980s, a began Orange, San Diego trend that accelerated immediately folThe 61 Loc Led Agenci(S( couny and three cty healh are listed under th soir repecive region. lowing the imposition of the 25-cent cigarette tax increase in January 1989. When the full campaign took effect To track changes in the health care field, the SDSU eval- one year later, this trend appears to have accelerated even a uators conducted a statewide assessment in 1990, 1991, and little further, indicating some combined effect of the media 1992 of physician and dentist practices regarding tobacco and LLA and CG interventions. cessation counseling, referrals, knowledge of tobacco control Whether this divergence can be attributed to the tax, resources, and availability of cessation-prevention materials the programs, or both, the results are even more impressive in their offices. The response rates were relatively low, rang- from the perspective of the percentage reduction in the curing from 20% to 35% and typical of other mailed surveys to rent levels of purchasing. health care providers. To provide additional information, Assumptions among public health and tobacco control spot checks of physicians' offices were conducted on the experts have always been that a hard core of the smoking responses regarding availability of materials. population will find it difficult to quit and that continued reduction in consumption at this point will be even more difThe management information system. Beginning in 1992, ficult. Nevertheless, California continues to cut into this supdata from LLAs, CGs, and the UCSD statewide telephone posed hard core at a rate superior to that experienced nationsurvey were periodically summarized and distributed to wide, where overall purchasing is much higher. In terms of LLAs and CGs in a newsletter format. This quarterly average quarterly percent decline, California showed an avernewsletter contained articles, tables, and graphs detailing age quarterly decline of 3.6% from 1980 through 1988 and a information about cessation, prevention, and policy-ordi- 7.9% average quarterly decline from 1989 through the last nance activities implemented by the various projects. quarter of 1994. The rest of the country showed a 2.4% To provide comparative feedback, each newsletter con- 1980-88 decline and 3.2% 1989-94 decline. tained a table with data from specific forms comparing each LLA and CG. Additionally, each LLA received an individ- Local lead agencies. LLAs implemented more than 10,000 ualized insert containing graphs with specific variables so multi-session prevention, cessation, and environmental the individual LLAs could visualize their own achieve- tobacco smoke (ETS) programs during this time period. ments. The comparative and individual feedback was Sixty-one percent of these programs focused on prevention, intended to provide both positive and negative reinforce- 37% on cessation, and 2% on ETS. In 1993 and 1994, proment for improving the quality and quantity of standardized grams focusing solely on ETS grew in number as the issue data reported to the evaluators. became widely and publicly debated. Programs focusing FMncisco, San-\ BerkeleyFrsoKa
July/August 1996
*
Volume
I
ai, KIngs, Madera Merced, Tulare
Public Health
Reports
355
Scientific Contribution
Figure 2. California and the rest of the United States: wholesale cigarette purchasing trends, adjusted per capita 50X..
..... . .....
..... .-............
.....
.......-.-.....
.....
....
.,...
....
.........
UnitedStates 40
.....
. ..
...........
....... ......
.~~~~...... W35.. 1
........
......I..
........
.......
......., .....
...
........
........
.........
.,,,,,,,,.
WX..+z|.@B@ ..... ,.
25.
35
..
.....
.......
......... *
.*.......nia. *.......
.........
20. *. .
25*..
20 *-
.......
15
,
.........Ai
...
*................
.
.....
v -*@ "
.....
........
.........I...................I........ ........*-,-
..........
80 81
..............
.......
....-.... .
.....
,
*
t
i
,
82
83
84
85
86
i
.
87 88 Year
i
,
,
,
.
*
90
91
92
93
94
emphasized traditional prevention and cessation topics. Toward the end of 1992, however, the sessions began covering such topics as advocacy, community resources, youth access, tobacco industry tactics, and environmental tobacco smoke. The greatest change in the types of activities LLAs were involved in, as more emphasis was placed on environmental and community level interventions, was in the area oflocal tobacco control policies and ordinances. Only 1% of activities were reported in this arena in 1990; more than 53% were reported in 1993. Technical assistance was the major activity in this area, largely because local health departments are not allowed to use these funds for political or lobbying efforts. Between 1990 and 1994, there was an increase in policies and ordinances activities relating to access to minors with a concurrent decrease in activities relating to dean
Reports
.........
...
89
training sessions primarily
indoor issues.
...
....... ...... ........... .......
.........
solely on smokeless tobacco remained few (approximately 3%) and even decreased somewhat in the first quarter of 1994. LLAs also conducted more than 2,000 training sessions in which organizations, professionals, young people, and a myriad of other community constituents were trained in the skills of tobacco control. In the majority of cases, these sessions focused solely on cigarettes, with very few emphasizing smokeless tobacco. Early in the program,
356 Public Health
.........$...............
......
95
Tobacco industry counterefforts were reported in the form of attendance at meetings and lobbying of local council members, distribution of anti-ordinance propaganda, advertising, and formation of front groups. Industry counterefforts were highest during 1991 when more than 17% were concentrated in policy-ordinance activities. LLA activities reached more Latinos than any other specified ethnic group, and this trend held throughout the four years. Across every type of activity, American Indians were targeted the least often, with programs reaching this group tapering off dramatically by the end of 1993.
Competitive grantees. CGs implemented more than 7,700 multi-session prevention, cessation, and ETS programs. Fewer than 2% of these programs focused solely on smokeless tobacco. As many as 25% of the programs were conducted in Spanish; 17% were conducted in a language other than Spanish or English (most often Chinese or Vietnamese). Grantees were active in the policy-ordinance arena, with activity peaking in 1992. Of the forms submitted, 44% reported providing technical assistance, 39% enactment of a new policy, and 17% reported revision of existing policies. The majority of policy activities focused on dean indoor air (52%); approximately 30% activities ~~~~~~of
related
by
to
access minors. In more dtan 21% ofthe
limiting
slightly
reports, counter-
tobacco industry efforts were confirmned; industry counter-efforts were suspected in 4% of the reported activities.
CGs conducted more than 1000
in
training sessions prevention, cessation,
ETS, tobacco industry tactics, community resources, and advocacy. Although the topic of prevention remained at the forefront throughout the four years, sessions regarding ETS, tobacco industry tactics, and policies increased significantly during 1993 and remained high during the first quarter of 1994. School-age youth were most likely to be trained by CGs, followed by current tobacco users, parents, pregnant women, and out-of-school youth. More than 22,000 Californians reported attending a CG session. As with the LLAs, Latinos were targeted most often (30%), followed by July/August 1996
*
Volume
Ill
California!s Tobacco Education
African Americans, Asians/Pacific Islanders, and American Indians. Regional linkage activities. All of the regions were involved in local media campaigns, with a total of 114 local campaigns reported. Three regions stood out with their use of local media campaigns: Gold Country, Los Angeles, and Tri-County together accounted for 75% of the reported campaigns, with 48% of the population. More than half (57%) of all the reported media campaigns targeted both rural and urban areas, while 32% targeted urban areas only, and 11% targeted rural areas only (Figure 3). A strong effort was made to reach ethnically diverse audiences, with almost an equal effort reported across ethnic groups. Although English was the primary language in the majority of campaigns, Spanish was used in more than 32% of the reported media activities. A total of 130 special events were conducted by the regions, with the Gold Country responsible for 51%. The regions involved a wide variety of local community groups in these activities. In a continuing effort to reach as many people as possible, the regions conducted more than 200 training sessions. Topics ranged from youth empowerment to ETS to smokeless tobacco to tobacco industry tactics. More than 5400 people attended these various sessions. A variety of ethnic groups participated; Latinos in 95% of the sessions, American Indians in 56%, African Americans in 54%, and Asians/ Pacific Islanders in 23%.
Tobacco industry response. In California, a different type of advertising has largely replaced traditional tobacco advertising in newspapers, especially in papers with less then 100,000 circulation. Much of the advertising was in the form of between-the-pages inserts. These inserts are often full page and announce events, refute anti-tobacco charges, and make so-called socially responsible claims such as, "We Figure 3. Geographic targets of regional media campaigns, California tobacco control
July/August
1996
*
Volume
I
think smoking is for adults, not minors.' A decline in outdoor tobacco advertising was observed from 1990 to 1993, but in 1994, that trend was reversed. In the first quarter of 1994, the average number of cigarette ads observed for the monitored areas was up 26%. The upswing in outdoor tobacco advertising paralleled an increase in outdoor advertising in general and may be related to a general change in marketing strategy. The California anti-smoking campaign has reached a level of intensity in minority-oriented newspapers nearly equivalent to the tobacco advertising in these papers. About as many anti-tobacco ads were observed in these papers as tobacco ads and promotions. The volume of cigarette ads were similar for Latino, Asian, and general audience newspapers in California-five to eight ads per 100 issues. However, the tobacco advertising rate was three times higher in California newspapers targeted to African Americans-15 ads per 100 issues. The most striking feature of tobacco promotions in California over the past four years has been the shift away from traditional advertising toward strategies designed to reach the potential smoker more directly and countermand local, state, and Federal tobacco control efforts. The most obvious of these new strategies is the upsurge in point-of-purchase or in-store promotional items. As tobacco advertising in traditional media outlets has decreased, in-store promotions in the form of displays, signs, give-aways and youth-oriented merchandise have increased significantly, more than doubling in California observation sites in the past two years. Initially, this strategy was carried out primarily in predominantly white neighborhoods; however, in the last year a faster rise in in-store promotions was observed in Asian neighborhoods. Finally, one of the fastest growing trends in tobacco marketing is the sponsorship of entertainment and commuPublic Health
Reports
357
Scientific Contribution
nity events. In California, the tobacco industry appeared particularly interested in sponsoring "growth events" whose participants are expected to grow rapidly over the next few years (for example, in Tejana music). Based on the events sampled in California, tobacco promotions reached nearly three million residents in Los Angeles and Orange counties, not including television or other media coverage of the events. Sponsorship of events tended to favor sporting venues (79%) whose attendees were inclined to be young (48% younger than 25) and Latino (28%).
Primary health care. A wide range of Proposition 99 activities involved primary care physicians, dentists, and their staff members. Twenty-five different local health departments and 10 competitive grantees trained nearly 1500 health care providers on tobacco intervention techniques. Additionally, more than 1000 prevention, cessation, and ETS programs were conducted within the health care system channel. In the data available on participants from these programs, 55% of smokers reported that a physician had counseled them to quit, 22% reported a dentist had counseled them to quit, and 33% received no such advice.6 In the fall of 1993, a validation study was conducted to verify certain responses to the 1992 survey of physicians and dentists. A random sample of physicians and dentist offices in San Francisco Bay Area and Los Angeles was selected for site visits. Two independent raters determined if there were no-smoking signs posted in waiting areas and if there were stop smoking brochures or other materials available in the waiting areas. None of the physicians had stop-smoking materials available in their waiting areas, although on the survey 69% reported that they did. Seventy percent of the physicians reported to have no-smoking signs in their waiting areas, while only 15% actually did. Ofthe dentist offices visited, 25% had no-smoking signs in spite of the fact that 60% reported that they did. In terms of stop smoking materials, there was 100% agreement between what was reported and observed. Unfortunately, no dentists reported having materials in their waiting rooms.
Discussion The independent process evaluation of Proposition 99 tobacco control efforts in California involved a loosely connected set of activities and impact variables. The evaluation showed that local health departments, community-based organizations, and other groups produced an exceedingly high volume of diverse tobacco control activities throughout the state and attempted to reach a variety of ethnic groups and at-risk populations. American Indians, as a target group, were the least likely to be targeted and reached by any program, and this trend appeared to worsen as time went on. The data also indicate that LLAs and CGs responded to the overall shift in their thinking about tobacco control and moved their interven358 Public Health Reports
tions from being individually focused to a broader, more community-oriented, environmental focus. A failing of the process evaluation was its inability to relate specific program efforts to local impact on tobacco use. The design of the process evaluation was not originally set up to link activities to direct impact, and this became a source of frustration to tobacco control staff members. Although the management information newsletter was implemented to provide feedback to local program staff members, ultimately neither the LLAs nor the CGs found the information useful in improving program efforts. Future evaluation efforts of this scale need to be able to link the "dose" of program efforts with tobacco use "response" much more closely, both on a local and global level. Our evaluation indicated that Proposition 99 accomplished much of what it originally set out to do: reduce tobacco prevalence, implement tobacco education programs among previously underserved groups, and heighten the awareness statewide of the dangers of environmental tobacco smoke. Since 1988, smoking has decreased in California by 28% (more than twice as rapidly as in the rest of the country), due both to the increase in price and also the unique and extensive local health education and media programs funded through this unusual mechanism.7 The dedicated tax is an effective approach to tobacco education and control, especially with the possibility that public health programs may be relegated to state block grants. Furthermore, future efforts in the area of tobacco control must take into consideration the considerable skills and resources the tobacco industry will bring to bear to minimize any damage to its image and profit, and Proposition 99 has done just that.
This research was funded by the California Department of Health Services. The authors are grateful to Dileep Bal and Mike Johnson for their assistance. References 1. Bal D, et al. Reducing tobacco consumption in California: development of a statewide anti-tobacco campaign. JAMA 1990;264: 1570-1574. 2. Pierce JP, Burns DM, Berry C. Reducing tobacco consumption in California: Proposition 99 seems to work, JAMA 1991;265:
1257-1258. 3. Elder JP, Kenney E. Independent evaluation of proposition 99-funded efforts to prevent and control tobacco use in California. Interim report. San Diego: San Diego State University, 1992. 4. Series of 8 Regionwide Forecasts 1990-2015. San Diego Association of Governments, January 1994. 5. Preventing tobacco use among young people: a report of the Surgeon General. Atlanta: Office on Smoking and Health, Public Health Service, 1994. 6. Crooks CE, et al. Tobacco control activities of primary care physicians in California. Eval Health Prof 1993 September;16:311-321. 7. Toward a tobacco free California: mastering the challenges,
1995-1997. California Department of Health Services, Tobacco Control Section, Sacramento, 1995.
July/August 1996 * Volume I I I
